Finlay Colville, senior analyst, Solarbuzz, San Francisco, CA USA
Next year looks set to see record revenues recorded by many equipment suppliers. For them, the only question is how long this can be sustained beyond the next 12 months, before equipment demand softens again.
Leading c-Si producers in China are investing heavily in new capacities along the value chain as part of their long-term strategic positioning. However, the near-term impact is likely to be strong bookings for preferred equipment suppliers and a potential oversupply of low-cost solar panels during the second half of 2011.
During 2009 and 2010, tier 1 photovoltaic (PV) producers embarked on strong rates of PV capacity expansion, with annual growth rates exceeding 50%. The majority of the equipment installed for this represented c-Si capacity added at low cost, located within China and Taiwan. While originally motivated by reaching capacities at specific stages along the c-Si value chain at the 1GW level, 2011 is now poised to see this extended to 2GW.
|Figure 1:Equipment spending and PV book-to-bill for tier 1 PV producers between 2008 and 2011. The book-to-bill ratio stabilizing close to parity during 2H 2010 provides an early indicator that equipment spending during 2011 will be similar to 2010 levels. (Source: Solarbuzz PV Equipment Quarterly.) *Midstream here includes c-Si (ingot to module) and thin-film panel.|
For some, vertical integration at the 2GW level is a must-have attribute. For others, such as pure-play solar cell makers in Taiwan or poly/wafer producers, the strategic focus remains on long-term contracts and OEM supply agreements within the value chain.
Strategic positioning guides capacity cxpansion plans for 2011
The key driver behind c-Si vertical integration has always been to obtain the lowest integrated cost structure, and to be prepared for the possible threat of any market-ready thin-film technologies. But this alone does not account for the sheer scale of expansion currently underway in China by c-Si tier 1 producers.
Another motivation is to plug gaps in the value chain, thereby addressing immediate concerns over raw material availability or pricing. Indeed, this problem is anticipated to become more acute during 1H 2011 as competition between industry neighbors — hitherto customers and suppliers to one another — becomes more pronounced.
But the underlying motive is purely strategic, driven by long-term market share aspirations and with the goal of becoming a top 3 player beyond 2011. Capital equipment investment today ensures that vertically integrated tier 1 producers are ideally prepared tomorrow for subsequent phases of strong industry growth.
Tracking capacity installed and production levels achieved
The back end of 2010 has been awash with announcements of capacity expansion (read recent supplier releases by Oerlikon Solar, Meyer Burger, Sharp, all from December). At first glance, adding up all the MW and GW across each of the value chain stages makes for interesting projections for nameplate capacities by end 2011. More pressing, however, are the implications for low-cost panel supply during 2011 and the demand for fab tooling placed on the PV equipment supply chain. Therefore, means of tracking capacity expansion phases and capacity conversion rates (production as a function of ramped manufacturing capacity) are key requirements now, with clear metrics needed to monitor progress on a quarterly (or monthly) basis.
Figure 2:Module production and capacity conversion rates for tier 1 PV producers between 2008 and 2011. The 2011 values are based upon recent expansion announcements from tier 1 producers, using capacity conversion rates observed during 2H 2010. (Source: Solarbuzz PV Equipment Quarterly.) *Illustration for 2011 here is based on Tier 1 announced expansion plans at December 2010.
The manufacturing equipment PV book-to-bill provides one such barometer to check capacity expansion rates. This tracks new equipment orders, tool deliveries and equipment backlog on a quarterly basis. The book-to-bill approach is further validated when confining the analysis to equipment ordered by tier 1 PV producers. Indeed, the PV book-to-bill for tier 1 midstream (ingot-to-module and thin-film) fab equipment has been posting ratios well above parity since the start of 2010, and this is projected to continue until at least Q2 2011. (Book-to-bill ratios above parity indicate that more orders are being received by the supply chain than tools being shipped to fabs). Further, a healthy equipment backlog has been accumulated during the past few quarters, with a considerable amount of tooling set to ship during 1H 2011.
New terminology adopted
Capacity conversion estimates for 2011 are less predictable however, as market demand next year may be subject to more uncertainty than many are willing to consider. This accounts for the widespread concerns related to potential overcapacity within the industry during 2H 2011. Recurring arguments are being voiced however by tier 1 vertically integrated proponents to justify capacity expansions in the short term, with much attention here afforded in marketing campaigns to terms such as “brand awareness” and “bankability.”
However, while this strategy may be sufficient to discount the incremental capacity coming on-line from tier 2 and tier 3 players next year, the competitive threat from like-minded China tier 1 vertical expansion peers may provide a much greater challenge. Slogans aside, tier 2 and tier 3 players (within any industry) have intrinsically less brand awareness and supply product to the market that is less bankable, regardless of the prevailing competitive environment.
Tier 1 expansion plans may be subject to revision
Throughout the recent Q3 2010 reporting season — and subsequent 2011 guidance — there has been a somewhat selective rationale considered by each tier 1 producer when projecting market share gains. However, rather than focusing on individual company’s share estimates, the more pertinent questions relate to the effect of the collective tier 1 plans. How many tier 1 panels are likely to be produced during 2011? How does this compare to demand scenarios being forecast? And looking further afield to c-Si cell makers in Europe, Southeast Asia and Japan with higher processing costs, how much will this increase the outsourcing currently being employed within the industry?
If tier 1 manufacturers enact upon their 2011 capacity expansion guidance — and simply maintain their collective capacity conversion (or utilization) rates throughout each quarter next year — tier 1 suppliers alone will produce some 21-22GW worth of PV panels. Remember, this does not factor in the 200+ tier 2 and 3 manufacturers, many of whom are intent on expanding capacity next year or are seeking to make the transition from pilot to low-volume production. And finally, add in an increasing number of c-Si module manufacturers in China rushing to integrate backwards with cell lines, fearing a repeat of cell shortages or simply concerned that their previous suppliers are positioning themselves as directly competitive at the module level. Excluding higher risked expansion announcements, tier 2 producers have the scope to produce a further 6-7GW of modules during 2011.
In practice however, various caveats must be placed upon capacity expansion and vertical integration aspirations. First, vertical integration targets are often qualified as being ‘flexible’ or ‘partial.’ This suggests a more pragmatic approach to capex spend, with greater risk in execution at value chain stages out with historical (technology) comfort zones. Also, short-term expansion is being guided by visibility of order books extending over 1H 2011. In this respect, manufacturing ramp-up phases during 2011 that are more back half weighted should be assigned with a higher risk.
High-efficiency c-Si Cells
It may be reasonable to assume that while demand is strong, most of the high-efficiency-enabled lines will run in standard process flows. And when oversupply takes effect, high-efficiency tooling will be incorporated more vigorously (assuming equipment has indeed been purchased, shipped, and qualified). For new high-efficiency c-Si advocates however, due regard should be accorded to the timelines from lab to mass production and the incremental capex on record from SunPower (back junction), Sanyo (heterojunction), BP-Solar (‘Saturn’ selective emitter), and Suntech (‘Pluto’ selective emitter). Or the success of companies such as Advent, Photovoltech, and Solland in bringing rear contact (wrap-through) c-Si cells to mass production along with their requirement for customized module assembly lines.